Article Text

Download PDFPDF

Self-monitoring of blood glucose did not improve glycaemic control in patients with type 2 diabetes not treated with insulin

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A Farmer

Dr A Farmer, University of Oxford, Oxford, UK; andrew.farmer{at}



randomised controlled trial (Diabetes Glycaemic Education and Monitoring [DiGEM] study).




blinded (laboratory staff).*



48 general practices in Oxfordshire and South Yorkshire, UK.


453 patients ⩾25 years of age at diagnosis of type 2 diabetes (mean age 66 y at study entry, 57% men, median duration of diabetes 3 y) who were managed with diet or oral hypoglycaemic agents alone, had haemoglobin A1c (HbA1c) level ⩾6.2% (mean 7.5%), and were independent in activities of daily living. Exclusion criteria included serious disease and use of a blood glucose meter ⩾2 times/week in the previous 3 months.


self-testing by use of a blood glucose meter 3 times daily 2 days/week, with instructions to contact their physician if readings were consistently high or low (self-testing group, n = 150); self-testing (as above) plus training in timing, interpreting, and using the test results (self-monitoring group, n = 151); or quarterly blood glucose tests but no use of a blood glucose meter (control group, n = 152). All patients received usual care, including training in behaviour-changing techniques and feedback on glycaemic control.


change in HbA1c level, blood pressure, cholesterol, weight, and body mass index; and hypoglycaemic episodes.

Follow-up period:

12 months.

Patient follow-up:

87% (intention-to-treat analysis).


At 12 months, groups did not differ for change in HbA1c level (table), blood pressure, weight, body mass index, or ratio of total cholesterol to high-density lipoprotein cholesterol. Decrease in total cholesterol level was greater in the self-monitoring group than in the control group. Risk of hypoglycaemia was highest in the self-monitoring group (table).


Blood glucose self-testing or self-monitoring did not improve glycaemic control in patients with type 2 diabetes not treated with insulin.

Self-testing with a blood glucose meter (ST) v self-monitoring (SM) v no use of a meter (control) in type 2 diabetes not treated with insulin*

*See glossary.

Abstract and commentary also appear in ACP Journal Club.


Farmer A, Wade A, Goyder E, et al. Impact of self monitoring of blood glucose in the management of patients with non-insulin treated diabetes: open parallel group randomised trial. BMJ 2007;335:132.

Clinical impact ratings: IM/Ambulatory care 7/7; Endocrine 6/7; GP/FP/Primary care 5/7


Some experts consider SMBG to be an essential part of managing all patients with diabetes, especially those using insulin. Among patients who use tablets or diet to treat their diabetes, the benefit of SMBG has remained uncertain. Universal use of SMBG in this group has important cost- and treatment-burden implications. A meta-analysis supported by industry suggested benefit,1 although a Cochrane review concluded that more evidence was needed.2 The UK National Health Service funded the DiGEM trial to answer this question definitively.

The trial was designed to detect a potential dose–response relation across 3 groups: no SMBG, less-intensive SMBG, and more-intensive SMBG (with accompanying education). Participants had reasonably well-controlled diabetes at baseline. Somewhat surprisingly, more participants in the more-intensive self-monitoring group stopped doing SMBG than in the less-intensive group. This finding may reflect lack of interest in SMBG among study participants: Patients already monitoring were excluded from participation, and they may have derived greater benefit from intensive SMBG. Furthermore, patients in the more-intensive group who persisted with SMBG did monitor more often, but whether they accrued important benefits from SBMG remains unclear. Overall, the results suggest that the benefits of SMBG in patients treated with tablets are minimal at best.

In addition to considering the findings of this trial, clinicians should listen to what their patients have to say about SMBG. A recently published qualitative study of SMBG highlights, among other insights, that clinicians’ focus on HbA1c levels when assessing diabetes control may be interpreted by patients as indicating that SMBG is not important.3


View Abstract


  • Source of funding: National Health Service and National Institute for Health Research Health Technology Assessment Program; Abbott Diabetes Care provided blood glucose meters.